This work is licensed under a Creative Commons License.
I grew up with a background of Physics, Mathematics and Engineering. That's where my understanding of graphs came from. Later I got a bit of a rough grounding in statistics, geometry and other useful bits and pieces. Graphs are such a great way to represent measured data and look for patterns, it's great to see them turning up in newspapers.
If we really want to solve the problems we say we do, barrow-pushing is no substitute for hard-headed analysis and the gathering of empirical evidence.
That's what we need -- hard-headed analysis. What makes some heads harder than others, I'm not quite sure but there can be no doubt whatsoever that hard-headed analysis is better than soft-headed analysis. I haven't bothered to draw my own version of the image, follow the link here or the article link above.
I'll quote from Ross Gittins to explain what the graph is all about:
The outside, heavy line is employers' potential demand for labour, while the inner dotted line is workers' potential supply of labour.
In search of deep understanding I pondered on this definition... at all points on the graph the outside, heavy line (demand) is representing a greater number of individuals than the inside dotted line (supply). Thus, we can conclude that for all wage brackets demand is greater than supply except for the highest wage brackets where demand is equal to supply. In the lowest income segments, the demand outstrips supply by a factor of 2:1 or more.
Hmmm, if demand outstrips supply in a jobs market then what would that make unemployment? I guess it would be negative. Australia has reached the rather exciting achievement of a negative unemployment rate, all thanks to the ingenious Ross Gittins and his incredible graph.
There are a number of other illogical conclusions that come from the above -- if demand is twice as high as supply for low wage earners then why are their wages so low? One would expect a free market to encourage employers to fill their great demand by offering better wages.
People falling into areas C and D are those whose marginal productivity exceeds the value to them of the welfare payments, leisure and other opportunities (including home production and production in the grey economy) they could gain by being outside the workforce.People falling into areas A and B are the opposite: their welfare, leisure and other opportunities are worth more than the market wage they could earn.
In the light of the supply/demand definition given above, it seems patently obvious to me that areas "A" and "B" cannot possibly contain unemployed people (because the graph already shows unemployment is negative). When demand is greater than supply, the gap between supply and demand is made up of unfulfilled demand, or in this case job vacancies waiting to be filled.
I'm certainly a subscriber to the belief that Economics is rightly known as "The Dismal Science". All the time I hear about how economists have to be experts in mathematics and then I look at what they actually come up with and (so far) every time I look closely, there are obvious flaws and highly questionable presumptions all over the place and if there is any real mathematics, it is only a veiled cover for complete rubbish underneath.
That's not to say I have given up on economics, but we have to be honest with ourselves and admit that current technology and analysis techniques are far, far away from actually being able to understand the subject and the results we get must be tempered with a healthy dose of skepticism. It would seem that there are enough Australians who do believe what the economists tell them to the point where economists have become the new soothsayers and witchdoctors of the modern age. They can't see the future any better than Joe the Plumber can, but they sure know how to put on a good show of mumbo-jumbo to keep the crowd enthralled.
In this case, I realised that I was being a bit hasty in my judgement. Ross gives credit for this graph to Paul Frijters and Robert Gregory, and their original article is available here, from Australian National University.
Taking the trouble to read the original article is quite enlightening, Pity Ross Gittins reported a different explanation for the same graph.
[The graph above] takes the 2005 distribution and divides the population of men into those for whom welfare and opportunities outside market work are worth more than the wages they could obtain in the market (i.e. their marginal productivity) and into those for whom the converse holds. Think of outside opportunities as an amalgam of many things, including production in the grey economy, leisure, and home production.
The key phrase here being "divides the population of men", so the outside, heavy line is actually supply -- that being the entire population of men (note that Frijters and Gregory focus primarily on men -- ignoring the rise in equal opportunity -- and also seem to disregard part-time or casual work). In fact, the whole graph is describing the supply side of the situation. We are looking at how much of the available supply of labour has actually been used up, and how much is still available that could be used under alternative conditions.
The whole demand side is missing from the graph; consider that in the real-world employment market there are always three distribution curves to consider:
Frijters and Gregory have a comment to attempt to explain this:
One main thesis of neo-classical economics is that in the long run everybody can have access to a job paying the marginal value of ones' skills. If a person gets paid less by the employer, some competitor will offer them a higher paying job which they will eventually find. If they get paid more by their employer, the employer would eventually go bankrupt. Though they may have trouble finding a job in the short run, in the long run one can either set up a job oneself or some entrepreneur looking to make a profit will create a suitable job. Of course this is a stylized and highly restrictive description of the main neo-classical thesis but this is nevertheless what many economists and policy makers believe must be true in the long-run. A key observation that supports this view is that labour force participation levels do not vary significantly with the level of population in a country, implying that in the long run it's the supply of labour that creates demand rather than anything else.
The problem with this conclusion, is that there is no meaningful "long-run" result in an industrial economy. Any data that we care to look at will be a snapshot, and in that snapshot will exist both employers demanding skills that are in short supply (and unable to fill vacancies) and workers offering skills that are not in great demand (and unable to get a job). The reality is that we live in a constant state of flux that will never reach a "long-run" equilibrium. This is especially true in the technology industry where requirements change rapidly as new technologies shake up the landscape of the market -- new options become available, old techniques become obsolete.
There are a vast number of variations of employee skills and jobs utilising those skills. Frijters and Gregory place those skills along a single axis of wages paid, which is a reasonable top-level summary but it hides a highly multi-dimensional mix of supply and demand in the job market, plus the effect of constant change in that market -- one day electricians are in short supply and can push up their fees, the next day everyone want carpenters, or programmers or whatever.
There's a number of other strange presumptions in the belief that "everybody can have access to a job paying the marginal value of ones' skills". How exactly do we measure the value of an employee's skills to a company? I would argue that no company with more than 10 employees has the ability to accurately make such a measurement. Moreover, an employee who makes a great improvement in a company's efficiency and then loses his job soon afterwards will continue returning a profit to the company long after they are gone but will get paid nothing for that. Contrariwise, a manager who makes a bad strategic decision and then gets a promotion and moves on to a different department will result in long term losses to the company but keeps getting paid. This is especially true when strategic decisions are made by a faceless committee and no one actually stakes their name and reputation on any particular outcomes -- the analysis of such decisions is usually only clear many years later, and sometimes no analysis is ever done as the people in question find ways to sweep their mistakes under the carpet.
The idea that companies that do it wrong will eventually go broke and this all will come good in the "long-run" is laughable -- companies go broke every day and new companies start just as rapidly. Other companies get bought, sold, merged, outsourced, tiered, liquidated, downsized and structurally adjusted. This happens constantly as one trend in management proves unworkable and another, more exciting bandwagon is rolled out. As I said earlier, technology is a never ending race as it keeps moving forward. Even when companies don't go broke, that doesn't prove they have an accurate measurement of all employees marginal productivity, it only proves they are not doing any worse than anyone else and that the total profit is sufficient to keep them afloat.
Other people argue, "the market may not be right, but it is the best estimate that we have got". This is a much more reasonable assertion, but if we take this assertion as true then the graph of Frijters and Gregory is impossible to draw for any real-world data because the market value of a potential wage earner who is currently out of work is an indeterminate figure. The only information that we can extract directly from the market itself is the current wages of the people who are working -- any other data must be constructed based on external presumptions.
Frijters and Gregory admit the lack of input data:
The truth is that we simple don't know, mainly because it is exceedingly hard to get a good empirical estimate of the productivity of individuals who don't have a job.
So the entire outside, heavy line in the graph is an imaginary estimate curve with a shape that could be pretty much anything depending on your point of view. Since the curve is a "potential" curve anyhow and we are talking about choosing policy that will cause deep changes to the economy, on what basis do we even say that the "potential" curve is stable? Is this the potential for every worker under ideal conditions to express their skills to the maximum effect (even when such conditions do not currently exist)? Or is this merely the potential for workers to find whatever job they can under current conditions? How far do we go as we keep imagining potential wages that may or may not exist?
Furthermore, Frijters and Gregory are happy to admit that there exist situations where the wage of a working employee does not equal their marginal productivity:
In a situation where technological advances require such re-allocation, the rigidity inherent in enterprise bargaining would prevent many profitable re-allocations from happening. Think only of the seniority system that operates in many firms; or the fact that demotions are only now becoming a legal possibility in many civil service jobs in Australia (the Australian National University has only in 2005 introduced the possibility of demotion).
Once we accept that wages are only a rough estimate of actual productivity we now have a gap between what people are paid and what they are actually worth and we are left with no easy way of even knowing how large the gap is or which direction it goes.
Consider this report mentioned by the ABC:
The report by the Australian National University shows the wealthiest 1 per cent of the population receives almost 10 per cent of the country's income.The report's author, Dr Andrew Leigh, says the annual income of a chief executive is more than 98 times that of an average worker.
I'm going to go out on a limb here and say that there is no possible way that a chief executive really does equivalent work to 98 of the average employees in any company. Even accounting for multipliers caused by a better organised workforce or higher productivity owing to better strategic planning, a factor of two orders of magnitude is just beyond belief. We can only conclude that political factors and positive feedback caused by the CEO's ability to influence their own remuneration come into play here. The CEO gets a large payment because they are highly skilled at ensuring they hold onto their position of power and can manipulate the system enough to prevent any competition for their job.
There's one last point where I find Frijters and Gregory's conclusions rather disturbing. Not because they are inaccurate as such but because they sidestep several key factors.
The issue of shifts in the distribution of marginal productivity raises the obvious issue of whether there isn't some way in which we could simply train everybody in areas A and C in order to get them into D? The answer is that various Australian governments have tried that for decades. In terms of more education per member of the population, this policy has been a resounding success: since 1970 the proportion going to university has more than doubled to about 40% of current cohorts; the proportion failing to finish secondary school has dropped to only about 25% (down from 80% in the 50s); and there's been a great expansion of other education programs, such as TAFE. It so far hasn't reversed the effects of skill-biased technological change. The poor effectiveness of training programs and Active Labour Market programs targeting the skills and incentives of the long-term unemployed have been well-documented for Austraila and other countries.
Nothing we do will ever reverse skill-biased technological change. You can't turn back the clock, end of story. Technology is a treadmill where you have to keep running just to stay in the same place. The "third world" are no longer as primitive as we thought they were and Australia went for a long time without modernising it's shipping, manufacturing, transport and other fundamental industrial segments. In the last 15 years, governments have steadily bled money out of tertiary education, teachers wages have been kept low and although more students sit in school, the strength of various qualifications have certainly been watered down. Thus, our education system has pushed us forward a bit but everyone else is moving forward faster than we are. That doesn't mean we should give up on education as a strategy, it just means that we have to put more into it to make it work.
Citing poor effectiveness of retraining the long-term unempolyed is only marginally relevant. Most important are the young unemployed who are not long term but are at risk of becoming that way. By far our first priority must be to provide top quality education for the kids and youth of Australia. This is an issue that has come up at the last two Federal elections and which the Australian voters have ignored by voting for John Howard and his education cuts so we should fully expect to be even further behind in our national skills base as this generation of Australians grow up.
Ensuring we get the most out of the children from disadvantaged backgrounds through high minimum education standards seems the obvious starting place for a policy reaction.
All education standards have to be kept high. Ignoring the average while concentrating on the minimum is nice for equality but nuts for future productivity. We have a huge challenge facing us -- surviving global warming with our food production intact, coping with the inevitable rise of disease as antibiotics become useless, finding alternative energy sources to replace oil, the list goes on. None of these problems are priorities at the moment, all of them require an excellent supply of educated people. Our marketplace economy has demonstrated an inability to plan ahead, it only reacts to problems after the problem is already serious and even then in an erratic and unreliable manner. We have to plan ahead ourselves rather than wait for the "invisible hand" to come along and sort things out.
In conclusion, although many policies are based off worse arguments than the A, B, C, D graph regions, Frijters and Gregory's graph is a pretty weak policy development tool. It ignores skills matching between education, population and industry; it is based on questionable presumptions and is impossible to refine into anything really empirical. The economic analysis of putting everything onto a single scale based on price is merely a way of hiding the physical world behind a smokescreen that keeps everyone ignorant. We also have to stop our blind trust in the market and start looking at real cause and effect so we can start measuring parameters that are meaningful.
This work is licensed under a Creative Commons License.